Self-play

Technique · Training · 2 items since Jul 2026

A model improves by competing against, or setting challenges for, copies of itself.

Trends

Oct 6
MedZERO self-evolves medical reasoning
A self-evolving framework pairs an examiner that generates frontier medical questions with a reasoner that solves them via evidence-grounded, tool-augmented multi-turn reasoning, improving 4B and 8B models by up to 13.7 accuracy points over prior self-evolving baselines on five medical benchmarks.
Jul 15
OpenAI's GPT-Red automated red-teamer
Trained by self-play RL, the red-teaming model finds prompt-injection attacks on 84% of held-out scenarios (versus 13% for human red-teamers) and was used to adversarially train GPT-5.6 Sol to roughly sixfold fewer failures on OpenAI's hardest direct-injection benchmark.